Conference Agenda

We propose a regression discontinuity design to determine whether

there is a difference in the expected returns of green versus brown secu-

rities (greenium). We analyze daily changes in option-implied expected

returns following the outcome of close vote ecology-related shareholder

proposals to estimate the causal impact of a small change in investors’

aggregate greenness perception on expected returns. Upon proposals’

passage, expected returns with a forecast horizon of 365 (730) days drop

by approximately 9.3% (6.9%) relative to proposals’ failure, indicating

the existence of a negative greenium in US equity markets. Our results

are robust to the introduction of a battery of fixed effects and when ac-

counting for multiple periods around shareholder proposal voting dates.

We propose a new approach for predicting the equity risk premium (ERP) that first estimates expected returns on individual stock before market-level aggregation. Our model combines firm-level return predictions from neural networks trained on a two-dimensional feature set of post-publication firm-level characteristics and aggregate macroeconomic variables. This method achieves an out-of-sample R² of 2.74% from 2000-2021, with strong economic significance in trading strategies after transaction costs. While the market generated a return of 376% over this period, a simple market-timing strategy based on our model's forecast signs yields a net cumulative return of 768%. Our results challenge the conventional wisdom that the ERP is unpredictable out-of-sample, suggesting that valuable market-wide information can be extracted from the cross-section of individual stocks.

We study the trading behavior of financial intermediaries around Federal Open Market Committee (FOMC) announcements in the S&P 500 index options market using intraday data. Proprietary trading firms are net sellers of options on FOMC days, in contrast to other days, with their trading activity concentrated in the morning, well before the announcement. Larger option sales by proprietary trading firms in the morning predict both a more accommodative monetary policy shock later in the day and a subsequent decline in option prices after the policy announcement, rendering morning trades profitable. We decompose monetary policy shocks into three components. Our analyses suggest that some financial institutions may hold an informational advantage regarding the Fed’s projected future policy actions.

This is the first study to analyze bankrupt firms’ reorganization plans. Using machine learning, we generate a dictionary for predicting post-bankruptcy success from these documents. Word counts based on our dictionary predict post-bankruptcy survival even after considering variables utilized in previous studies. Our text-based metrics are the strongest predictors of firm survival in our analysis and are also informative about the operating performance of surviving firms. Our results highlight the potential of reorganization plans for predicting post-bankruptcy success. We demonstrate that established dictionaries mostly evaluate reorganization plans incorrectly, which emphasizes the need for context-specific dictionaries

Motivated by recent technological advances and the importance of corporate culture for organizational transformation, we investigate the role of tech open corporate culture in a firm's capital market performance during technological disruptions. We develop and validate a measure of corporate tech openness culture based on Glassdoor employee reviews. Using the launch of a major large language model as our setting, we show that tech openness is positively related to firms' stock market performance following disruptive technological advances. Consistent with corporate culture driving the effect, our baseline results are stronger when investors are more aware of firms' culture. These findings suggest that capital markets recognize tech openness as a valuable intangible asset, highlighting the importance of human factors in technological transformation.

Tax havens are used for tax minimization. Whether tax havens affect corporate control in the form of cross-border mergers and acquisitions (M&A) or are merely used as conduits between host and destination countries of (greenfield) foreign direct investment and portfolio investment is an open question. We provide new stylized facts through the first comprehensive analysis of cross-border, tax-haven mergers and acquisitions (M&A). Using novel tax residence data, we investigate 18,193 such transactions from 1990 to 2017, totaling $7.1 trillion in deal value, or 31% of cross border M&A volume. $4.1 of the $7.1 trillion exceeds our prediction based on a gravity model with economic fundamentals. Small havens such as Bermuda alone make up $2.3 trillion or 10% of cross-border M&A volume.

This paper explores how peer effects influence homeowners' decisions to insure against elemental damages. Using Facebook friendship links to flooded areas, it finds that stronger social ties increase insurance purchases in non-flooded regions: a doubling of social ties leads to a 1.2% rise in policies covering elemental damages. It highlights the moderating role of regional climate policy attitudes and social capital, showing that bridging (bonding) capital weakens (strengthens) the effect. This contributes to understanding how social capital and climate attitudes shape household responses to climate risks.

This paper empirically analyzes the drivers of corporate greenwashing behavior to enhance the forecasting and mitigation of greenwashing practices, particularly in the context of stakeholder decision-making. Using company-level characteristics of a sample of STOXX Europe 600 constituents, we show that ESG ratings have a U-shaped relationship with greenwashing, indicating that greenwashing is more likely for companies with low and high ESG ratings. Additionally, environmental score, company size, cash-to-assets, and capital intensity are positively correlated with greenwashing. Furthermore, greenwashing behavior is more prevalent in consumer-related industries than in other industries. After identifying the determinants of greenwashing, we investigate the forecasting of greenwashing behavior using machine learning models based on economic considerations.

This pre-registered online experiment examines how environmental impact uncertainty influences asset allocation decisions. Our experiment precisely controls the environmental impact risk by linking investment decisions to the retiring of CO2 emission certificates. Results show that most participants adjust their investment strategies when confronted with impact risk, especially when such risks affect financially safe assets. Highly risk-averse investors show the strongest reduction of investments in impact-risky assets. We provide causal evidence that such modifications in investment behavior stem from changes in risk perceptions. These findings suggest the need for enhanced risk assessment frameworks that integrate environmental impact uncertainty in financial decision-making.

This paper analyzes probabilities of default (PD) and book leverages of firms. Both are almost uncorrelated. The development of PDs is modelled as an AR(5) process. We find mean diversion of PDs in the short-run and mean reversion to long-term target PDs over longer time intervals. The expected PD does not converge monotonically to the target PD, but overshoots and oscillates with declining amplitude. The same is true of the book leverage. Including leverage in the PD regression does not improve explanatory power. The PD converges faster to the target PD starting at a high PD. The target PD of a firm is lower when more than one bank reports a PD, also if an owner bears unlimited liability.

We investigate the impact of taxes on bank liquidity creation using the Tokyo bank tax as a quasi-natural experiment. This tax is on the gross profit of large commercial banks operating in the Tokyo prefecture. Using bank-only taxes offers better identification than prior research using taxes covering banks and non-financial firms. We find the tax significantly decreases liquidity creation, consistent with empirical domination of the Risk Absorption Hypothesis over the Financial Fragility-Crowding Out Hypothesis. Banks shift from long-term loans liquefying the public and supporting the economy to short-term government securities holdings that decrease public liquidity without economic stimulus. The tax also impairs bank capital and risk management, making the financial system riskier.

How does risk perception in credit markets change after observing a nearby catastrophic event? We combine detailed geospatial data on ex-ante flood risk of German firms with credit register data and show that after a major flood in 2021, loan rates decrease for high-flood risk firms that were not directly affected. This negative indirect effect is strongest for banks with a large loan portfolio exposure to the flood. Firms that were affected by earlier, but similar floods do not experience rate reductions. The decrease is also strongest in areas with low climate change belief, while high climate change belief areas experience rate increases.

I extend the robust mean-variance portfolio analysis proposed by Maccheroni et al. (2013) by examining how ambiguity prudence affects optimal stock allocation when the investor evaluates portfolios using the smooth ambiguity model. Ambiguity prudence reflects an aversion to model uncertainty that intensifies as the investor believes unfavorable events as more likely. I derive a higher-order approximation of the certainty equivalent to disentangle the contributions of preferences and beliefs in payoff valuation. Ambiguity prudence introduces nonlinearities into the investor’s valuation, leading to sizable deviations from the robust mean-variance solution, and contributing to information inertia when the investor exhibits greater concern about downside model uncertainty.

This paper analyzes optimal compensation contracts when managers are overconfident.

We separate the two components of overconfidence: optimism (overestimation of ex-

pected firm value) and miscalibration (underestimation of the firm value’s volatility).

We calibrate a stylized principal-agent model to the each of the observed contracts of

3.370 CEOs from 2008 to 2021 to obtain the optimism and miscalibration measures. In

our empirical study, we find that CEO miscalibration is correlated with leverage, debt

and equity issue, whereas CEO optimism is correlated with R&D expenditures.

This study investigates how S&P 500 CEOs use interviews after earnings announcements to influence market perceptions, as their role in managing reactions to earnings surprises remains underexplored. Across 1,869 announcements during 2023–2024, we find that the probability of a CEO holding an interview rises with the magnitude of the surprise, particularly negative ones. These interviews help mitigate short-term price declines but correlate with weaker future firm performance. Interviews mirroring earnings calls boost returns more than those adding new insights, our LLM/topic-model analysis shows. Also, CEOs pivot to global themes after negative surprises but stress firm details after positive ones, reflecting expectation management. Overall, these findings suggest that interviews sway short-term sentiment rather than align it with long-term managerial expectations.

We compare all commonly employed transaction cost measures in options markets. Standard measures do not clearly identify the financial crisis, suggest that low-volume ITM options are much more liquid than high-volume ATM options, and are highly sensitive to small moneyness variations. We propose normalizing the relative bid-ask spread by the option elasticity, effectively capturing the direct cost of implementing the replicating portfolio and the option's economic exposure. This elasticity-adjusted spread outperforms alternative measures. It varies with market stress and equity market liquidity and shows high cross-sectional correlations with underlying liquidity, market capitalization, and volatility. It is lowest for ATM and ITM options and robust to small variations in moneyness. Approximation of our measure based on daily data performs similarly well.

We examine the variance risk premium (VRP) in carbon markets. On average, the VRP is positive but turns negative during periods of market stress, such as the COVID-19 pandemic and the Russian invasion of Ukraine. We decompose the VRP into upside (“good”) and downside (“bad”) components. We find the VRP to predict short-term carbon returns on specific carbon events. Importantly, the predictive power for future carbon returns is entirely driven by the downside VRP, indicating that investors are particularly sensitive to jump risks linked to falling carbon prices. This sensitivity likely reflects broader economic concerns, as declining carbon prices typically indicate reduced investor demand driven by lower emissions, diminished industrial activity, and a weakening economy.

We examine the positioning of different types of traders with respect to option return anomalies. Anomaly demand varies widely across variables and between call and put options. With their anomaly exposures, customers are overwhelmingly on the wrong side of anomalies. Market makers are the main beneficiaries. Preferred habitats explain the exposures to multiple anomalies, while end-user demand also emerges as a likely driver of several anomalies. Aggregate end-user trading patterns suggest that they mainly seek outright rather than delta-hedged option exposure. Finally, market makers appear to dynamically adjust prices based on firm trader demand, but not customer demand.

We develop a continuous-time endowment economy model of the US with inflation and the central bank’s interest rate adjustments as observable risk factors. We show that they have predictive power for consumption growth and can explain many features of the aggregate stock and bond market. We derive the price-dividend ratio, the equity premium, the risk-free rate, and the term structure of interest rates. We show in a calibrated model that inflation and the federal funds rate adequately predict those key asset pricing moments. The model offers a novel mechanism to explain the variation in the aggregate price-dividend ratio and the risk-free rate as it relies on observable rather than latent risk factors.

This paper applies a two-stage evaluation procedure to study investors’ portfolio selection and asset pricing consequences in the presence of risk and ambiguity. The two-stage evaluation procedure represents investors’ attitudes towards risk and ambiguity. Investors optimally hold a combination of three funds: a risk-free, a solely risky, and a risky and ambiguous fund. In equilibrium, the risk premia on all assets are determined by their exposure to systemic risk and ambiguity. The results are straightforward to implement and test and well-suited for many types of applications. The results theoretically explain the asset allocation puzzle and the size effect.

We analyze an intermediary asset pricing model with three heterogeneous agents — funds, banks, and households — in a continuous-time production economy. Agents differ in risk aversion and capital productivity; financial intermediaries face a Value-at-Risk constraint. We show that the constraint significantly dampens investment, asset prices, and consumption growth. Recursive preferences amplify these effects, while log utility substantially underestimates them. Our results highlight the macroeconomic consequences of financial regulation and investor heterogeneity, offering insights for both policy design and the asset pricing literature.

I find that corporate insiders benefited from a temporary reduction in algorithmic trading during the SEC Tick Size Pilot Program. Insider purchase quantities increased by approximately 30% during the program without compromising quality, and reversed once the program ended. Using algorithmic trading proxies from the SEC Market Information Data Analytics System, I show that the effects were concentrated in stocks with above-median pre-pilot algorithmic trading activity. These findings highlight that microstructure shocks exert material impacts of informed insider trading.

We investigate how financial news media coverage causally affects managers’ manipulation decisions in corporate disclosures. While prior research shows that media coverage decreases manipulation incentives for “hard manipulation”, such as accounting fraud (“watchdog role”), the role of the media for manipulating qualitative information, “soft manipulation”, is ex-ante unclear; theory even predicts that news media may uncritically disseminate corporate information (“mouthpiece role”), thereby increasing managers’ incentives for soft manipulation. Using a stacked difference-in-differences design based on restructuring events at the Wall Street Journal, we provide first empirical evidence that news media coverage reduces managers’ incentives for soft manipulation. The effect is driven by articles with more editorial content and more pronounced for firms with less monitoring through other corporate governance mechanisms.

This paper examines whether responsible institutional investors affect corporate green capital expenditures. I identify responsible investors via the Climate Action 100+ initiative and measure green capital expenditures with green debt issuance. Given regional differences in investor sustainability preferences, I exploit the cross-listing of European and Asian firms in the United States as negative responsible ownership shock. A staggered difference-in-differences estimation shows that cross-listing firms have lower responsible ownership and decrease green capital expenditures post cross-listing. Lower responsible ownership is also associated with lower green governance. Together, my findings suggest that responsible investors drive green capital expenditures by influencing corporate decision-making.

This paper analyses the impact of banks' dividend restrictions on the behavior of banks' institutional investors. Using within-investor variation and a difference-in-differences setup, I find that mutual funds, especially high dividend-paying funds, reduce their ownership in treated banks during the 2020 payout restrictions in the Eurozone. This reduction is not reversed after the abrogation of the policy. Using data before the introduction of the ban also reveals a positive relationship between fund ownership and banks' dividend yield, highlighting the importance of dividends for European banks’ fund investors. This reaction also has pricing implications as suggested by a negative relationship between cumulative abnormal returns around the restriction announcements and the percentage of fund owners per bank.

We construct a global capital buffer announcement dataset, capturing the direction and intensity of measures for the pandemic years 2020 and 2021. Using a panel VAR model, we examine the responses of stock prices to these announcements for 52 countries. We find that stock prices reacted negatively to capital buffer easing announcements. The negative response was driven by Capital Conservation Buffer (CCoB) announcements, while markets barely reacted to announcements related to other buffers. We find no differences between the responses of overall and financial sector indices. Moreover, sample splits reveal cross-country heterogeneity in stock price responses to CCoB announcements. Our recommendations support a stronger reliance on the Countercyclical Capital Buffer (CCyB) during periods of stress.

We study the impact of state-contingent dividend taxes (and bans) and capital regulation on a firm's optimal strategy and value. The firm generates stochastic income under time-varying macroeconomic conditions. Its manager distributes dividends and issues costly equity to maximize shareholder value. We solve the manager's stochastic control problem and derive the firm's reserve distribution in closed form. Imposing dividend taxes (or bans) during crises generates a trade-off, as it encourages reserve accumulation in bad states but promotes payouts in good ones. The policy undermines financial stability by reducing the firm's value and recapitalization incentives across states. Coordinating dividend taxes with counter-cyclical capital regulation can mitigate value losses and ameliorate the trade-off, but it also creates additional recapitalization disincentives.

Numerous studies show that stocks that attract a lot of investor attention tend to have low subsequent returns. Motivated by these findings, I proxy the relative visibility of stocks in the cross section with their ranked abnormal return volatility. I argue that this simple measure is consistent with salience theory as well as the concept of information gaps, curiosity, and motivated attention. For a sample of US large-cap stocks, I show that both high (low) relative visibility alone and combined with high uncertainty, as proxied by idiosyncratic volatility, lead to statistically and economically low (high) subsequent returns. The results are consistent with time-varying expected returns and a large predictable return component.

A substantial fraction of IPOs is priced at exactly the high-end of the pricing range. We find that these IPO prices are predictably biased. Investing in IPOs priced at the high-end leads to first-day returns that are 8 percentage points higher than investing in IPOs that are priced just below or above the high-end. We argue that these results are in line with managerial satisficing, in which managers stop seeking higher outcomes and settle for satisfactory results.

We study a dynamic communication model with a myopically loss averse investor as information recipient. In a multi-period setting, the manager learns about the firm's fundamental value earlier than the investor and can report this information either truthfully or with a bias. We assume that the manager's communication strategy aims at optimizing the investor's perception of firm performance. Our model predicts that the manager will try to avoid downward price movements, which are disproportionally detrimental to the investor. In particular, the manager will claim stock values close to the investor's prior expectation to avoid immediate or future down movements. We examine the asset pricing implications of this communication strategy and find that managerial behavior can reduce volatility and cause momentum.

We evaluate the pricing performance of a robust stochastic discount factor spanned by a large set of cross-sectional factor returns. Methodologically, we employ kernel principal component analysis to construct factors that are nonlinear functions of a high-dimensional set of firm characteristics. We show that combining nonlinearities with novel regularization techniques enhances model performance, increasing the out-of-sample Sharpe ratio of the efficient mean-variance portfolio by roughly 30\% to 2.25 and reducing pricing errors across a wide range of prominent cross-sectional stock return anomalies. Moreover, we find that the improved out-of-sample performance of the nonlinear model correlates with measures that capture the time-varying pricing complexity of equity returns.

Firms may enact competitive sorting to value-optimizing transition or status quo opportunities. If firms Roy (1951) sort, then the information arrival of a firm’s competitive strategy should result in a positive equity price reaction. We apply latent variable techniques from Heckman, Stixrud, Urzua (2006) using a novel dataset of active manager self-edits of ESG scores. We find positive 18-83 basis point stock reactions in industrial base firms when information arrives that firms competitively sort toward transition or status quo growth strategies. Our effects largely go away in countries with high environmental stringency (much of Europe and East Asia), suggestive of unobservability in our data or a pooling toward transition investment induced by policy.

We study the factor structure of currency characteristics by employing a three-dimensional tensor factor model that simultaneously captures the variation in characteristics of the G10 currencies over time. We show that factor-mimicking portfolios derived from these common factors in currency characteristics are able to price individual currency returns better than standard factor models derived from univariate sorts on the same characteristics. The variation in currency characteristics can be well captured by a parsimonious two-factor model, where the first factor closely resembles the carry trade and the second factor acts as a hedge against carry crash risk, that is composed of signals from FX momentum, FX value and the term spread.

Tokenization of real-world assets (RWAs) promises enhanced accessibility and tradability in traditionally illiquid markets. Yet, achieving sustained liquidity and efficient price discovery in digital asset markets remains uncertain. We examine three liquidity mechanisms—automated market makers (AMMs), peer-to-peer (P2P) marketplaces, and centralized buybacks—using a dataset of 444,535 tokenized RWA transactions (2019–2024). We document arbitrage-driven liquidity flows, market fragmentation, and variations in investor sophistication. AMMs provide continuous liquidity but face systematic arbitrage-induced drainage. P2P marketplaces facilitate efficient price discovery, while buybacks offer stable yet inflexible exits. Our findings highlight the necessity of hybrid liquidity models integrating centralized and decentralized features; otherwise, RWA markets risk evolving into fragmented digital search markets rather than efficient trading ecosystems.

Using novel transaction and quote data, we show that transaction costs in the SOFR swap market are 1) low on average; 2) higher for client-to-dealer than interdealer trades reflecting differences in (permanent) price impact; 3) increase in trade size (measured in terms of interest rate risk), are lower for package transactions, and increase in interest rate volatility, client order imbalance, and dealer credit spreads; 4) moderately higher than for similar-sized interdealer Treasury trades; 5) stable over time, except during the banking crisis in March 2023 and the tariff-induced crisis in April 2025. Overall we determine that the SOFR swap market operates efficiently, though the functioning of the client-to-dealer segment may deteriorate somewhat during crisis periods.

This study examines the economics of latency arbitrage on fragmented European equity markets using nanosecond-level message data. I identify and trace the lifecycle of 277,654 stale quotes on thirteen regulated exchanges for a sample of 624 equities. Findings reveal that latency arbitrage is neither instantaneous nor frictionless, as only 54 percent of arbitrage opportunities are exploited by High-Frequency Traders (HFTs). Thus, market makers are not always adversely selected when quoting in fragmented markets. Throughout the sample period, HFTs generate substantial profits of 1.4 million Euro at cost borne by liquidity providers. Heterogeneous market structures are the main determinant of empirically observable stale quote sniping patterns while HFTs rely on nonlinear variable combinations to facilitate final sniping decisions.

The growing share of intermittent renewable infeed calls for deeper integration of electricity markets. For Central Western Europe (CWE), day-ahead prices across bidding zones are often aligned, shaped by transmission capacities implicitly allocated through Flow-Based Market Coupling (FBMC). However, most established electricity price models overlook multi-zone coupling and lack implementable solutions to incorporate FBMC. We develop a multi-market model that embeds FBMC into a residual demand framework to forecast day-ahead electricity prices over a mid-term horizon. Using structural data on interconnectors and flow constraints, our model captures key market interactions across zones. This approach improves price forecasting and supports policy and investment decisions in an increasingly interconnected power system.

This paper enhances the implied cost of capital (ICC) estimation by improving analysts' multi-horizon earnings forecasts with machine learning, which demonstrates reduced bias and higher forecast accuracy. The resulting machine learning model-based ICC is approximately 17% lower than the traditional analyst-based ICC with optimization bias. The model-based ICC shows stronger cross-sectional relationships with future realized returns, thus providing a better proxy for ex-ante expected returns. In addition, stocks with underestimated ICCs exhibit significant outperformance relative to those with overestimated ICCs. Weaker information environments primarily drive the ICC differences between analyst-based and model-based ICCs. Moreover, greater ICC differences are associated with lower earnings announcement returns.

We find a consistent momentum return premium of 1.25% p.m. in the global market for luxury watches between June 2017 and September 2024. Consistent with the frog-in-the-pan hypothesis, investors are inattentive to a series of frequent gradual returns during the momentum formation period and draw more attention to infrequent large returns. Momentum portfolio profits decrease from 1.67\% for luxury watches with continuous information during their formation period to -0.38\% for watches with discrete information but similar cumulative formation-period returns. Conditional on continuous information, momentum profits covary with the market state, return dispersion, and investor sentiment. Our results indicate that the disposition effect is not responsible for the return predictability of continuous information.

This paper investigates the effects of offering sustainable investment options through a unique natural experiment that allowed individuals to switch to a plan with a strong sustainability focus. Using pension fund data combined with Danish population registers, we analyse the impact on individual consumption behaviours, particularly vehicle purchases and emissions. Offering sustainable pension options promotes sustainable consumption, primarily among individuals who decline the sustainable plan. Women, married individuals, and those with sustainable consumption patterns, such as owning electric vehicles, are more likely to select sustainable pensions, while higher income and financial sophistication predict lower uptake. These results highlight the indirect behavioural effects of sustainable investment choices and underscore the role of pension policies in fostering sustainable practices and societal change.

Bitcoin Ordinals and inscriptions facilitate the on-chain storage of arbitrary data on the Bitcoin blockchain, expanding the range of Bitcoin-related use cases. In this study, we analyze the impact of inscriptions on the Bitcoin network. We find that inscriptions significantly increased network activity, created additional demand for blockspace, and influenced Bitcoin’s fee market dynamics. Furthermore, we find that the rise of inscriptions coincided with an increased utilization of Taproot, a notable increase in block size, and the longest sustained period of full blocks in Bitcoin’s history. Our study shows that inscriptions have reshaped how Bitcoin's blockchain is utilized and underscores the growing number of use cases beyond its original function as a peer-to-peer financial network.

Using stock-level data from the securities lending market, we investigate how financial markets process information. Our key empirical innovation is to use demand increases in the securities lending market to pin down the arrival of information. A premise that we motivate theoretically and validate empirically using state-of-the-art metrics for informed trading. We find that, when a stock experiences an increase in its shorting demand, the return volatility and its correlation with trading volume increase while the price informativeness decreases. These findings align with a differences-of-opinion model where overconfident investors trade so aggressively on their perceived increases in information that prices reveal less about fundamentals. Our results remain after excluding public announcements, highlighting the securities lending market’s role in private information.

This article evaluates the impact of the European Central Bank's (ECB) third covered bond purchase program (CBPP3) on yields, issuance and portfolio positions using a matched difference-in-difference estimation. The CBPP3 is estimated to reduce yields for euro area issuers, relative to non-euro area issuers. In contrast, it appears to have stimulated issuance in both regions. From a demand-side perspective, bank investors exhibit the least elastic demand; an increase in their ex-ante investor share is associated with greater yield and issuance impact. Euro area banks buy and hold covered bonds relatively inelastically, as they can pledge them as collateral in ECB repos to obtain cash.

The Bitcoin network imposes significant external costs, including CO₂ emissions and electronic waste, comparable to those of entire nations. Some studies argue these externalities are justifiable, claiming that cryptocurrency mining fosters sustainable energy by monetizing surplus power. We examine this trade-off from an economic perspective, asking: Does surplus energy use mitigate Bitcoin’s externalities? Are carbon credits effective? And does mining ultimately benefit or harm society? Our model shows that while surplus energy reduces CO₂ emissions, it also incentivizes more mining, increasing e-waste and potentially offsetting gains. We find that carbon credits, if implemented multilaterally, effectively lower total externalities. Our study concludes by highlighting the dual role of Bitcoin mining in facilitating the short-term energy transition while hindering long-term decarbonization.

We develop and evaluate a training intervention to foster internet banking adoption in a field experiment with more than 25,000 elderly customers of a large German bank, of whom we randomize 333 into training. We find that the share of internet banking users increases by 26 percentage points relative to a matched control group. In terms of sustainable usage, the share of online transactions increases by 14 percentage points and remains elevated four months after the intervention. Our data allows us to estimate intent-to-treat effects and predict drop-out along the multi-stage adoption process, shedding light on practical roll-out considerations. We show that the type of training (self-guided versus social training) impacts drop-out differentially despite similar treatment effects.

This paper investigates how the introduction of Large Language Models (LLMs) has shaped academic writing in finance journals. Analyzing over 40,000 finance articles from 34 different journals, we find that readability declined while the use of LLM-associated terms increased following ChatGPT’s release at the end of 2022. These trends are more pronounced among authors from non-English-speaking countries and in lower-ranked journals. Moreover, the effects are strongest in countries that place less emphasis on creativity and technology and exhibit higher levels of moral confusion. Author-level analyses suggest that while LLM adoption boosts publication quantity, it does not enhance publication quality or scholarly impact.

We estimate the expected returns of machine learning-based anomaly trading strategies and quantify the impact of three factors often overlooked in the previous literature: transaction costs, post-publication decay, and the post-decimalization era of high liquidity. Despite a cumulative performance reduction averaging about 57% when accounting for these three factors, sophisticated machine learning strategies remain profitable, particularly those employing Long Short-Term Memory (LSTM) models. We estimate that our most effective strategy, the one based on an LSTM model with one hidden layer, has an expected gross (net) Sharpe Ratio of 0.94 (0.84). We rationalize these findings in a simple theoretical framework in which technological diffusion gradually erodes trading profits while superior signal processing capabilities allow the extraction of alpha from increasingly complex information.

Dealers intermediate trades in OTC markets through trading networks. In the municipal bond market, we document greater complexity than the typical core-periphery structure. Analyzing dealer reciprocity-the tendency to repay favors-we find reciprocity generally reduces markups. Dealers trade lower markups today for future liquidity. However, in small trading communities, reciprocity can foster collusion via quid-pro-quo agreements, inflating transaction chain markups. Among high-centrality dealers in large communities, high reciprocity lowers average markups by 80 basis points, while among low-centrality dealers in small communities, it raises markups by 72 basis points. Although only around 2% of transaction chains suggest collusive behavior, these significantly affect regression results, highlighting the importance of controlling for such outliers to accurately estimate centrality premiums or reciprocity discounts.

This paper analyzes the effects of hidden liquidity by studying the introduction of iceberg orders at a large cryptocurrency exchange. Compared to other assets, cryptocurrencies often trade against both fiat currencies and pegged stablecoins. Considering the introduction of iceberg orders for trading pairs against the US dollar but not against a dollar-pegged stablecoin, this study finds that hidden liquidity is associated with increased quoting and trading activity. Larger average trade sizes suggest greater institutional participation. Quoted liquidity improves while the price impact of trades declines. Realized spreads increase, indicating improved revenues for market makers. Price discovery also shifts significantly toward the markets accepting iceberg orders. Overall, our results suggest that hidden liquidity has positive effects on market quality.

We show that, while corporate bond ETFs systematically exhibit lower liquidity risk than the bonds they hold, they also face heightened intermediary risk. This effect is more pronounced for high-yield ETFs, for those with less liquid portfolios, and for funds reliant on weaker Authorized Participants. A stylized model reveals how partial segmentation between ETF and bond markets drives these diverging exposures. Overall, investors of corporate bond ETFs effectively trade reduced liquidity risk for increased intermediary risk, highlighting a fundamental trade-off embedded in the ETF structure.

We investigate whether mutual fund families favor their ESG funds at the expense of their non-ESG siblings. We find that the net-of-style return spread of ESG compared to non-ESG funds within the fund family is significantly greater than the gap with non-ESG matches outside the family. The difference is around 2% per year, on average, indicating sizable cross-fund subsidization that is mainly used to avoid underperformance of ESG funds. We link this difference in performance to various fund and family characteristics and relate the observed effects to measures of environmental awareness and fund flows. Additionally, we

investigate potential mechanisms of ESG favoritism.

This paper explores the trading dynamics of key participants in the German government bond market from 2005 to 2024, with a particular focus on hedge funds. Using a unique data set of the German Finance Agency we analyze trading flows across investor groups. The dataset allows us to decompose the group of foreign investors, who hold over 50% of German sovereign bonds, and identifies hedge funds as a pivotal group. Hedge funds have increasingly acted as quasi-market makers, absorbing newly issued bonds and providing liquidity. They assumed this role when quantitative easing started in 2015 and continued it during quantitative tightening, starting in 2022.

We study the drivers of the level, slope, and curvature of the term structure of European carbon futures. We find that commercial hedging open interest predicts changes in the carbon term structure. Speculative positions held by financial investors---which have grown in recent years---and alternative measures of financial speculation have no predictive value. Additionally, the ratio of natural gas to coal prices predicts changes in the term structure level, while crude oil and equity returns correlate with future changes in its slope and curvature. Our results suggest that futures prices of European Emission Allowances are primarily driven by hedging demand from firms covered under the EU Emissions Trading Scheme, whereas financial speculation plays a comparatively minor role.

We revisit the established dominance of local factors in international asset pricing by examining the predictability anomalies through the lens of factor momentum signals. Differing from previous results for asset pricing models where local factors are typically seen as superior to regional and global factors, our research reveals that regional and global signals surpass local signals in forecasting factor risk premiums. Investment strategies formed on non-local signals generally outperform strategies formed on local signals in all considered metrics. This outperformance is more pronounced for factors and regions with high structural integration. Non-local factor predictability potentially improves leading asset pricing models. Moreover, non-local signals revive momentum investing in markets previously thought to lack momentum opportunities, such as Japan.

The growing complexity of forecasting models increases the number of decision nodes in the research process, raising the risk of overfitting to specific design choices. We illustrate this issue using the recent concept of “pockets of predictability,” which posits that return predictability is time-varying and that short windows of high predictability can be identified ex-ante. In this study, we reassess the robustness and practical applicability of this approach. By analyzing 19,440 variations of the original methodology, we find that its effectiveness depends critically on various seemingly minor methodological decisions. Furthermore, return predictability has declined significantly in recent decades, and the potential economic gains are highly sensitive to trading costs. Overall, strategies based on pockets of predictability should be approached with caution.

Returns of OTM-index-options are negative on average and do not match expectations when only the observed equity risk premium is considered. Some studies argue that either jump or volatility risk premiums provide a sufficient explanation for the puzzle. We investigate replicability of earlier conflicting results and cover new aspects, including the impact of volatility, an alternative model for jumps, and the term structure of returns, based on weekly options.

Model predictions for the impact of volatility match evidence for equity options and alternative jumps do not enhance the explanation. With constant relative risk aversion adopted from previous literature, newer data and weekly options enable rejection of both jump risk and volatility risk premiums for the cross section of returns.

We decompose stock idiosyncratic volatility into long-run and short-run components and find that both are negatively related to delta-hedged option returns. The effects of the long-run and short-run components are explained by the limits-of-arbitrage and stock return jumps, respectively. Unlike the long-run component, the short-run component can be used to create a trading strategy that remains profitable after considering transaction costs. In downturns, only the short-run idiosyncratic volatility effect is significant. Further analysis shows that the limits-of-arbitrage’s explaining power arises from its intercept and common component, while jump’s explaining power arises from its residual component relating to corporate news arrivals.

We investigate how option traders take sides on the predictability of the underlying stocks, using a wide array of cross-sectional return predictors. To infer the aggregate trading direction of option end users, we look at cross-sectional differences in the prices of option portfolios replicating the stock positions, relative to the prices of the underlying stocks themselves. We find that option traders take a profitable side on signals associated with momentum, profitability, and volatility. In contrast, they take a non-profitable side on signals that are associated with investment, and valuation.

Major cryptocurrencies exhibit long-term declines in price growth rates and variance. This paper replicates these trends by appealing to the vast uncertainty about cryptocurrencies' eventual adoption demand, and a gradual learning process that reduces the uncertainty. In the model presented, the uncertainty leads to a discounted cryptocurrency price. However, investors receive a stream of noisy signals indicating the likely extent of eventual adoption, which reduces the discount over time. The model can forecast expected price growth and variance conditional on low interim adoption. The model also estimates the probability distribution of a cryptocurrency's extent of eventual adoption.

An easy-to-interpret rule for the optimal risk-adjusted social cost of carbon is derived using perturbation analysis. This rule internalises the adverse effects of global warming on the risk of recurring climate-related disasters and the risk of irreversible climate tipping points as well as the usual adverse effect on total factor productivity. It approximates the true numerical optimum well, especially if the small parameters (i.e., the share of damages in GDP, the sensitivity of the risk of disasters to temperature and the risk of climate tipping) are small enough and the discount rates corrected for growth

and risk is not too small. The rule is also accurate if applied to models with a different supply side.

This paper explores the economic drivers behind the widely used asset pricing factors SMB, HML, RMW, and CMA, suggested by Fama and French. While effective in practice, the fundamental economic origins of these factors remain vastly unclear. We analyze days with large factor returns and classify them using news articles from the following day. We find that HML is linked to macroeconomic news, while CMA is tied to news about commodities. SMB correlates with exchange rate news and a sentiment-driven factor, while RMW is influenced by firm-specific news. Our findings suggest that both risk-based and behavioral channels are at play, contributing to the debate on whether factor risk premiums are driven by rational explanations or mispricing.

Using a life-cycle model of household consumption and investment under family transitions, we investigate the long-run financial outcomes of divorces. We find divorce leads to a reduction in wealth, and the loss is larger for higher-educated women. Earlier fertility and a smaller parenthood penalty for women with only a high school diploma result in negative effects of divorces fading away by age 45, whereas for college-educated women, the same is achieved a decade later because of later fertility and a stronger parenthood penalty. Reduced economies of scale, switching to a single-person income, and losing wealth protection within marriage have the strongest impact on the divorced household economy.

We examine how financial advisors shape the elicitation of clients’ investment risk tolerance. Using data from 7,180 in-person meetings, we find that risk tolerance varies systematically across advisors. Clients assigned to male advisors exhibit significantly higher risk tolerance than their demographic twins advised by female advisors. These differences affect product recommendations and client portfolios, as clients follow advice despite advisor-induced noise. The pattern is not driven by strategic manipulation but aligns with a false consensus effect—advisors projecting their preferences onto clients. Our findings highlight the persistent influence of advisor identity, even in regulated environments, raising important questions for policy and practice on the objectivity of risk profiling and the behavioral roots of investment advice.

Previous research suggests that many investors view dividends as free income. In this paper, we show that this fallacy can drive investors to play with the house money. First, using brokerage data, we document that investors use disproportionate fractions of dividends to buy lottery stocks and options. Second, using an experiment, we replicate the tendency to gamble with dividends in the lab and find that this tendency is stronger among participants suffering from the free dividends fallacy.

Higher exposure to competition – measured by product fluidity – is associated with higher carbon emissions. This result is robust to using instrumental variables to obtain exogenous variation in fluidity and also holds when using only reported emission data, excluding estimated emissions. The relationship between competition and carbon emissions is stronger for firms in areas less concerned about climate change and areas with weaker social norms. Short-termism does not explain the results, as the emissions-competition link is at least as strong for firms with longer-term-oriented shareholders. Our findings suggest that promoting competition may be at odds with climate change abatement.

Using hand-collected data on European auto asset-backed securities (Auto ABS), we examine the role of mutual funds in financing the transition to zero-emission mobility. Mutual funds, particularly those with a green mandate, tend to have a higher exposure to sustainability-transparent Auto ABS and tend to allocate more capital to deals with a higher proportion of electric vehicles. However, we find no clear preference for portfolios with lower average CO2 emissions. This behaviour suggests that, in the absence of a framework for green securitizations, investors rely on sustainability proxies that are associated with the lowest disclosure processing costs. Our analysis provides important new evidence on how standardized sustainability disclosures at both the prospectus and loan levels could influence asset allocation.

Insurance companies can facilitate the net-zero transition by restricting insurance coverage for fossil fuel projects. We study the adoption, structure, and effects of carbon underwriting policies by the world's largest insurance groups. While such policies have become more prevalent and restrictive over time, their design and enforcement remain highly heterogeneous. Leveraging new data on U.S. coal mine insurance, we document that coal policies reduce the number of insured coal mines by 17%, insured coal volumes by 58%, and increase insurer-mine relationship terminations by 20pp. Nonetheless, some insurers with coal policies expand their coal underwriting.

The higher aggregate prevalence of loan over bond funding in Europe is not only driven by the well-documented differences in financial market settings but also strongly shaped by different firm characteristics. This paper shows that this debt choice depends foremost on firms‘ size and collateral availability. The European economy is much more fragmented than the U.S. economy, and thus features a different firm distribution. I estimate that if all European firms had access to a financial market like the U.S. market, their aggregate bond funding share would remain significantly smaller. This counterfactual suggests a limited potential for European corporate bond markets in the short and medium term.

Mutual fund families are increasingly assigning traders to manage corporate bond mutual funds. Using this setting to study the role of traders in investment management, we document that trader managers identify and exploit short-term trading opportunities at lower transaction costs. These skills are particularly valuable during periods of market stress. Moreover, trader managers exhibit sophisticated risk management behavior: They reduce credit risk during periods of market stress and take more maturity risk during periods of large interest rate fluctuations, while holding portfolios with greater convexity. The combination of these skills produces relative outperformance during periods of large interest rate fluctuations.

In a market-based financial system, the interplay between banking and capital markets is profound. Using comprehensive credit registry data matched with bank- and firm-level information from a small open economy, we show that enhancing access to international capital markets amplifies the procyclicality of credit supply and increases its exposure to global financial conditions. Although access to international capital markets boosts liquidity, its positive effects on the real economy are modest or negligible.

This paper exploits individual trading records from a large brokerage service to investigate the trading patterns of retail investors who take short positions in stocks. Short sellers' research activity does not suggest that they increase the amount of attention paid to stocks before taking short positions. Although short sellers report lower risk tolerance in their MiFID II questionnaires, their order behavior in short positions indicates greater risk-seeking behavior. Short positions, compared to long positions, constitute larger portions of overall portfolios and are more highly leveraged. Compared with other retail investors, short sellers perform worse, and their profit variability is greater.

Investment mistakes among self-directed retail investors are well-documented. Our study, based on data from a leading German private bank, provides new insights into High Net Worth Individuals (HNWIs). Their investment decisions are less tied to short-term needs and often supported by dedicated advisors. We explore whether HNWIs’ investment decisions are less biased than those of retail clients and how financial advice affects these biases. Our findings show that HNWIs display less overtrading, broader diversification, and lower susceptibility to the disposition effect. Financial advisors significantly improve diversification and risk management but have limited success in reducing overtrading among persistently active clients.

We show that the pricing of earnings information differs when market ambiguity is high. Using a sample of 616 S&P 500 firms, spanning from 2006 through 2023, we find ambiguity-induced pricing asymmetries between positive and negative news. Moreover, idiosyncratic risk is priced when the ambiguity is high. This supports the theoretical results of the Multiple Priors Utility. Ambiguity aversion impedes full diversification of idiosyncratic risk as portfolios are priced at their worst-case realization.

In this paper, we study the emergence and elimination of arbitrage opportunities in computerized limit order markets. We measure price changes, buy/sell market/limit order submissions/cancellations at message frequency during arbitrage opportunities in NYMEX WTI futures and their options and find that market makers update quotes before trades become profitable so that actual arbitrage trades hardly ever occur. In the more liquid futures market, executed trades eliminate arbitrage, whereas in the options markets, liquidity providers are the ones who correct arbitrage opportunities. These results highlight the critical role of market makers in maintaining price efficiency.

This study introduces a new approach for inferring the trade direction in option auctions by incorporating minimum price improvement requirements. Using a matched dataset of intraday transactions, this study shows that trades executed at a one-cent quoted spread involving fewer than 50 contracts, which are subject to a minimum price improvement of one cent, systematically buy at the bid and sell at the ask. The commonly applied quote rule misclassifies 95% of these trades. Accounting for this pattern increases the overall classification accuracy from 62% to 76%, and for options with less than one week to expiration from 56% to 81%.

Using data on 25 major U.S. macroeconomic announcements, I study how forecast dispersion affects market reactions to news. I find that the price impact of announcement surprises is negatively related to forecast dispersion, challenging the conventional view of dispersion as a good proxy for uncertainty. Instead, dispersion also reflects forecaster-specific noise. To rationalize these findings, I extend the endogenous information acquisition model of Benamar, Foucault and Vega (2021) to incorporate noisy forecasts. The model predicts that higher noise-driven dispersion incentivizes investors to acquire more pre-announcement information, increasing price informativeness ahead of the release and reducing the announcement’s price impact. My findings reveal an important channel through which forecaster heterogeneity shapes market reactions to macroeconomic news.

We empirically test Gabaix and Maggiori (2015)'s prediction that currencies are repriced by the country's external capital dependence when financial constraints of FX intermediaries change. Using solvency indicators, we develop a novel intermediary constraints index capturing risk-bearing capacity. We find that constraints are a priced risk factor in currency portfolios sorted by countries' net foreign assets. Portfolios of external debtors (creditors) have higher (lower) intermediary risk premia, but pay lower (higher) returns when constraints tighten. Tightening constraints are associated with a depreciation of countries with low net foreign assets, particularly emerging markets with high net debt and low FX reserves.

While Federal Reserve swap lines are now the primary policy tool for easing offshore U.S. dollar borrowing cost during times of stress, its pass-through to the FX market remains poorly understood, largely due to a lack of data on the OTC FX swap market. Using a bespoke settlement dataset, I conduct the first comprehensive empirical study of agent positioning in FX swaps around swap-line take-ups globally across jurisdictions, currencies and time. I show that swap lines lower U.S. dollar borrowing cost not only through a reduction in non-U.S. bank demand for U.S. dollar (substitution channel), as is commonly thought, but also through an increase in U.S. dollar supply (arbitrage channel). My results offer novel policy implications.

We examine violations of uncovered interest parity (UIP) in high frequency, accounting for the discrete nature of interest rate payments in foreign exchange markets. Exploiting both regression- and portfolio-based tests, we do not reject UIP during overnight trading but strongly do so during the U.S. intraday period. Furthermore, we document a strong divergence in excess returns for currency trading strategies exploiting UIP violations on announcement versus non-announcement days. The cross-sectional carry strategy earns the bulk of its excess returns on macro and FOMC days, whereas the dollar carry strategy generates positive returns on non-announcement days and depreciates on announcement days.

We use satellite-based production facility location data and provide evidence that firms from the cement and steel industry with production facilities within the EU leak emissions to countries not affected by the EU Emission Trading Systems (EU ETS). They do so by emitting more in their non-EU facilities and by acquiring new facilities outside the EU. Emissions are more likely to be leaked to pollution havens. Additionally, firms headquartered in countries with more developed financial markets and in civil law countries as well as private firms leak more emissions. Financially constrained firms are more likely to leak emissions within their existing network of facilities, while unconstrained firms are more likely to acquire new non-EU facilities.

This paper examines the causal effect of venture capital (VC) on scientists’ selection into entrepreneurship, using the 1979 reform of the prudent-man rule under the Employee Retirement Income Security Act (ERISA) as a natural experiment. By relaxing pension fund allocation restrictions, the reform significantly expanded the capital available to VC firms. I construct a novel historical dataset of U.S. scientists from the 1960s and track their business formation activities. Exploiting variation in reliance on tangible versus intangible capital across specialties, I find that scientists’ business formations doubled following ERISA. The effects are stronger among those with intangible specialties and those employed in the private sector. These scientists were productive inventors. I propose two mechanisms.

This study examines the relationship between stock market fluctuations and mental health-related hospitalizations in Denmark, leveraging detailed administrative data. We find that U.S. market returns exert a significantly greater influence on Danish hospital admissions than local or other international markets, highlighting the critical role of global market dynamics in shaping stress-related health outcomes. Our analysis further identifies employees of publicly listed firms and stockholders—particularly those with substantial ownership—as more vulnerable to market volatility.

This paper examines the transmission of investor attention between firms based on geographical proximity and product market similarity using EDGAR search traffic data. We find that firms receive more attention when a nearby firm experiences abnormal attention, with the effect decreasing with distance and over time. The effect is stronger for firms in similar product markets. Effects are most concentrate among nearby firms, with attention increases exceeding 90% for very similar markets, dropping to 41% as distance increases. This research extends the literature on investor attention and local information bias, highlighting the importance of peer firm relationships in

information transmission.

We use prediction market data from Polymarket to assess firm-level sensitivity to

the 2024 U.S. presidential election. Election sensitivity correlates with firm characteris-

tics, such as tax burden, China exposure, environmental performance, and interest rate

risk, and predicts ex-post returns. Using price contribution analysis, we show these

characteristics are priced at varying speeds, creating temporary inefficiencies. Variable

importance analysis reveals that, while election sensitivity is the superior predictor, its

marginal contribution is modest, reflecting its overlap with firm-level covariates.

This study examines how firms choose between debt and leasing to finance productive assets, particularly following the introduction of IFRS 16. By treating leases as equivalent to debt on financial statements, IFRS 16 provides a quasi-natural experiment to explore this decision-making process. Our findings reveal that firms increased their leasing activities post-IFRS 16, driven by managerial incentives tied to EBIT and EBITDA, while simultaneously reducing direct capital expenditures and credit quotas. Notably, firms with higher pre-existing leasing volumes and stronger EBITDA-based incentives exhibit the most pronounced shift. Contrary to concerns about managerial opportunism, our results suggest that this reallocation enhances investment efficiency and aligns managerial incentives with long-term firm value.

We show that individuals consider counterfactual outcomes when evaluating their investment decisions. Using individual investors trading-data, we find evidence of a Relative Disposition Effect: The likelihood of a sale is higher for stocks that performed better than an alternative investment than for stocks that performed worse than the alternative investment. The Relative Disposition Effect exists when considering the overall market as well as the focal stock's industry as alternative investment. It is distinct from (and can even subsume) the original Disposition Effect, the Portfolio-Driven Disposition Effect, the Rank Effect, and cannot be explained by stock characteristics. The evidence thus highlights that individuals do not evaluate financial investments in isolation, but in relation to alternative investments they could have made instead.

This study investigates the performance of loans that are granted to small and medium-sized enterprises underlying asset-backed securities in several European Union countries before and after the 2018 EU Securitization Regulation. Using loan-level data from European DataWarehouse covering the period from 2013 to 2023, matched with the originating banks' financial data and macroeconomic indicators, we find that the Regulation is significantly associated with improved quality of securitized SME loans. The impact is stronger for deals with low-performing originating banks in terms of capital adequacy, asset quality, liquidity, and profitability. We find no significant effect of loan replenishment post implementation. Conditional upon the post-securitization period, the Simple, Transparent, and Standardized label appears to be weakly associated with better loan performance.

In this paper, we use state-of-the-art machine learning models to analyze whether ESG scores, or their components, have predictive power for stock prices. Our results show that neither ESG scores nor their sub-scores improve predictive performance of machine learning models when models can select from an extensive set of financial predictors used in the asset pricing literature. This finding suggests that the information of ESG scores is either reflected in stock prices or that commercial ESG scores and sub-scores are noisy measures of firms’ ESG quality and thus have no predictive power of stock prices.

This paper examines whether mergers in the tech sector increase the co-movement of tech firm returns with the broader market. Drawing parallels to banks, we argue that tech firms can become systemically important by engaging in mergers. Using a stacked DID approach on 1,319 NASDAQ tech firms (2010–2023), we measure return co-movement using risk commonality measures. On average, we find no effect of mergers on risk commonality. However, large mergers raise risk commonality in periods of low merger activity, and rival firms also exhibit increased risk commonality during merger waves, suggesting a free-rider effect.

This paper investigates the impact of options trading on price volatility and hedging effectiveness in grain futures markets by analyzing the 1936 ban on commodity options trading in the United States. Utilizing newly collected data for wheat and corn futures from Chicago and London (1934-1939), we employ a difference-in-differences approach to compare treated (Chicago) and control (London) markets. Our findings show that active options trading is associated with reduced price volatility and improved hedging effectiveness in futures markets. Following the 1936 ban, Chicago futures experienced a significant but temporary short-term increase in volatility and a decline in hedging effectiveness, highlighting the important role of options trading in risk management, stabilizing futures markets and facilitating efficient information transmission.

Online platforms such as CloudResearch, Amazon MTurk, or Prolific require researchers

to communicate average expected payoffs to participants prior to experiments. We show

that knowledge of expected payoffs introduce confounding effects in experimental studies

on risk-taking. We use the literature that shows that individuals with lower ranks take

higher risks as a playground. Our experimental design disentangles this ranking effect

from a reference point effect introduced by the average expected payoff. Holding the rank

constant, risk-taking is 18.23% higher below the reference point on average. Holding the

reference point distance constant, ranks have no significant effect on risk-taking. These

results are robust to nonsocial settings and alternative risk-taking measures.

Using unique data from LinkedIn profiles, we measure the adoption of AI technologies by mutual fund managers. Compared to low-AI funds, high-AI funds generate superior returns and incur lower expenses. AI outperformance is particularly strong among discretionary funds, which rely on human judgment. The greater the AI adoption, the more pronounced the time-varying skill of fund

managers across different market conditions. The stock-picking abilities of high-AI funds improve with the availability of big data, such as satellite imagery of parking lots. The local availability of AI skills is a key determinant of cross-sectional variation in mutual fund AI investment. Our findings are robust to using geographic variation in AI supply as an instrument for AI utilization by mutual funds.

We evaluate the performance implications of providing various types of domain-specific information to large language models (LLMs) in a simple portfolio allocation task. We compare the recommendations of seven state-of-the-art LLMs in various experimental conditions against a benchmark of professional financial advisors. Our main result is that the provision of domain-specific information does not unambiguously improve recommendations. We find that LLM recommendations underperform recommendations by human financial advisors in the baseline condition. Providing firm-specific information improves historical performance in LLM portfolios and closes the gap to human advisors. Performance improvements are achieved through higher exposure to market risk and not through an increase in mean-variance efficiency within the risky portfolio share. Risk-averse investors are recommended substantially riskier portfolios when firm-specific information is provided.